Fig 2: Inhibition of PARylation down-regulates AMPk pathway. (a, b) Effect of PARylation inhibition on ATP levels and AMPk activation. Treatment with PJ34 and PARP-1 silencing (60 nM) strongly delays starvation-induced loss of ATP levels. Right panel shows the efficiency of PARP-1 silencing.*P<0.05, **P<0.01 versus the indicated group in MCF7 GFPLC3 cells by t-test (a). MCF7 GFPLC3 cells treated with PJ34 (left panel) or silenced for PARP-1 (right panel) during starvation showed down-regulation of AMPk activation (b). (c) Effect of PAR accumulation on ATP levels. Accumulation of PAR polymer hadn't effect on starvation-induced loss of ATP in starved MCF7 GFPLC3. The efficiency of PARG knock-down was previously shown in Figure 1a.**P<0.01 comparing between control and starved MCF7 GFPLC3 cells and øøP<0.01 comparing between control and starved PARG silenced MCF7 GFPLC3 cells. (d) Activation of AMPk in absence of PARG. The accumulation of PAR polymer in MCF7 GFPLC3 cells increased the activation of AMPK under nutrient deprivation. (e) AMPK activation is delayed in absence of PARP-1. Cell extracts of parp-1+/+and parp-1-/ -MEFs after different times of starvation were used to measure the levels of AMPk activation. (f) AICAR reactivated the AMPk/mTORC1/LC3 pathway in absence of PARP-1. Cell extracts from parp-1+/+, parp-1-/- and parp-1-/- MEFs pre-treated with AICAR 1 mM were used to measure the levels of AMPk activation (phospho-AMPK), mTORC1 activation (phospho-p70S6k) and autophagy induction (endogenous LC3-II translocation). (g) Starvation-induced autophagy is increased in parp-1-/-MEFS after AICAR treatment or PARP-1 reconstitution. *P<0.05, **P<0.01 versus the indicated group by t-test. In B, D, E and F figures total AMPk and total p70s6 kinase were used to normalize for the non-phosphorylated protein and ß-actin or GAPDH were used as loading control. Similar results were obtained in three independent experiments

Fig 3: Membrane binding of hLKB1 is essential for its function in mammalian cells.(a) Alignment of DmLKB1 and hLKB1 reveals several positively charged aa in the C terminus of hLKB1. Mutated residues are marked with bold lines. (b,c) GFP-hLKB1 localizes predominately to the nucleus and plasma membrane of HeLa cells cotransfected with STRADa (b), whereas GFP-hLKB1?LB C430A is not recruited to the cortex (c). (d,e) Immunoblottings of lysates from transfected HeLa cells, which lack substantial LKB1 expression, demonstrate an increase of activated AMPK (phospho-T172 AMPK), activated MARK (phospho-MARK) and decreased mTOR activation (measured by phosphorylation of S6K) upon GFP-hLKB1+STRADa transfection, which is not observed in GFP-hLKB1?LB C430A+STRADa transfected cells (d). Co-transfection of PLD2 but not of a catalytically reduced variant (PLDY511F) together with hLKB1 results in a further increase of AMPK activation (e). The intensity of pAMPK bands was quantified (normalized against total AMPK) (e, lower panel, control was set as 100%). (f) Cell viability of HeLa cells under energetic stress (induced by incubation with AICAR for 12 h) was estimated using the MTT assay (as described in the methods section). Scale bars are 10 µm. Experiments were performed in triplicates. Error bars represent s.d. and statistical significance was determined using ANOVA: P<0.01, **P<0.05, not significant (NS).

Fig 4: Mechanical loading activates mTORC1 to prevent sclerostin expression in osteocytes. (a) Representative images of the stretch-loaded MLO-Y4 cells for 0, 12 and 24 h. The scale bar represents 100 µm. 0 h: no-loaded group/control group; 12 h: stretch-loaded for 12 h; 24 h: stretch-loaded for 24 h. (b) Western blot analysis of SOST, pS6 and pS6K expression in MLO-Y4 cells stretch-loaded for 0, 12 or 24 h. (c) mRNA levels of SOST in stretch-loaded MLO-Y4 cells (one-way ANOVA with Bonferroni multiple comparison, 0 h versus 12 h, p = 0.0077; 0 h versus 24 h, p = 0.0003, n = 6). (d,e) MLO-Y4 cells were stretch-loaded and treated with 1 nM rapamycin for 24 h, then cell lysates were subjected to (d) western blotting or (e) qPCR analysis for SOST (non-parametric statistical test, vehicle group, p = 0.0246; ?R group, p = 0.0473, n = 6). All experiments were repeated independently three times. NL, no-loaded; SL, stretch-loaded. Data are represented as mean ± s.d., *p < 0.05, **p < 0.01 and ***p < 0.001.

Fig 5: mTORC1 is required for Akt induction of NF-?B in HNSCCA. Overexpression of Akt increases activity of mTORC1 and NF-?B. Different doses of active HA tagged Akt2 were transfected in Cal27 cells and phosphorylation p65 and S6K and total p65 and S6K, as well as HA-Akt2 expression were tested by western blot. B. Inhibition of Akt, mTORC1 and NF-?B by Akt inhibitor, perifosine. Cells were treated with different doses of perifosine for 24 hours and the levels of p65, Akt, S6K and of endogenous phosphorylation of S6K, p65, and Akt were determined by probing with the indicated antibodies. C. Depletion of Raptor blocks Akt induction of NF-?B. Cells were transfected with non-target siRNA or siRNA against Raptor as indicated for 48 hours and active Akt2 were transfected for another 24 hours. Cells were lysed and the levels of p65, Akt, S6K and of endogenous phosphorylation of S6K, p65, Akt2, and Raptor were determined. The results are representative of three experimental repetitions.